1. Field of the Invention
The present invention relates to transport of Internet Protocol Version 6 (IPv6) packets by IPv6 nodes via an IPv4 network having a Network Address Translator (NAT) or a Port Address Translator (PAT).
2. Description of the Related Art
Proposals have been made by Internet Engineering Task Force (IETF) groups for improved mobility support of Internet Protocol (IP) based mobile devices (e.g., laptops, IP phones, personal digital assistants, etc.) in an effort to provide continuous Internet Protocol (IP) based connectivity. The IETF has two working groups focusing on mobile networks, a Mobile Ad-hoc Networks (MANET) Working Group that is working to develop standardized MANET routing specification(s) for adoption by the IETF, and NEMO (mobile networks). NEMO uses Mobile IP (MIP) to provide connectivity between mobile networks and the infrastructure (e.g.,, the Internet). The key component in NEMO is a mobile router that handles MIP on behalf of the mobile networks that it serves.
A “Mobile IPv6” protocol is disclosed in an Internet Draft by Johnson et al., entitled “Mobility Support in IPv6”, available on the World Wide Web at the IETF website address “ietf.org/internet-drafts/draft-ietf-mobileip-ipv6-21 .txt” (the disclosure of which is incorporated in its entirety herein by reference). According to Johnson et al., the Mobile IPv6 protocol enables a mobile node to move from one link to another without changing the mobile node's IP address. In particular, the mobile node is assigned a “home address”. The “home address” is an IP address assigned to the mobile node within its home subnet prefix on its home link. While a mobile node is at home, packets addressed to its home address are routed to the mobile node's home link, using conventional Internet routing mechanisms.
The mobile node also is assigned a home agent for registering any care-of address used by the mobile node at its point of attachment to the Internet while the mobile node is away from its home link. A care-of address is an IP address associated with a mobile node that has the subnet prefix of a particular link away from its home link (i.e., a foreign link). A home agent is a router on a mobile node's home link with which the mobile node has registered its current care-of address. While the mobile node is away from its home link, the home agent intercepts packets on the home link destined to the mobile node's home address; the home agent encapsulates the packets, and tunnels the packets to the mobile node's registered care-of address.
A variation of the Mobile IPv6 protocol is disclosed in an IETF Internet Draft by Soliman et al., “Hierarchical Mobile IPv6 mobility management (HMIPv6)” October 2002, available on the World Wide Web at the IETF website address “ietf.org/internet-drafts/draft-ietf-mobileip-hmipv6-07 .txt” and incorporated in its entirety herein by reference. The Internet Draft by Soliman et al. discloses a Mobility Anchor Point (MAP) within an IPv6 network that implements HMIPv6 by assigning a regional care-of address to mobile nodes within its address realm. Mobile nodes may thus use on-link care-of addresses for communications within the address realm of the MAP, and the regional care-of address for IPv6 communications outside the MAP address realm. As such, the MAP serves as a local home agent.
Hence, a mobile node is always addressable by its “home address”: packets may be routed to the mobile node using this address regardless of the mobile node's current point of attachment to the Internet. The mobile node also may continue to communicate with other nodes (stationary or mobile) after moving to a new link. The movement of a mobile node away from its home link is thus transparent to transport and higher-layer protocols and applications.
Proposals are underway by the Next Generation Transition (NGTRANS) Working Group of the Internet Engineering Task Force (IETF) to enable network nodes to transmit IP packets, generated according to IPv6 protocol as specified by the Request for Comments (RFC) 2460, across an IPv4 network. In particular, RFC 3056 proposes an interim solution (referred to herein as “the 6 to 4 proposal”) of sending IPv6 packets as payload for IPv4 packets, where an interim unique IPv6 address prefix is assigned to any node that has at least one globally unique IPv4 address. These RFCs are available at the IETF website on the World Wide Web at “ietf.org” and these RFCs 2460 and 3056 are incorporated in their entirety herein by reference.
The 6to4 proposal specifies that an IPv6 node has an IPv6 address that contains an assigned IPv4 address, resulting in an automatic mapping between the IPv6 and IPv4 addresses. Hence, the IPv6 node can easily encapsulate the IPv6 packet with an IPv4 header based on extracting the assigned IPv4 address from within its IPv6 address. In particular, RFC 3056 specifies that the Internet Assigned Numbers Authority (IANA) has permanently assigned one 13-bit IPv6 Top Level Aggregator (TLA) identifier under the IPv6 Format Prefix “001” for the 6to4 proposal: the numeric value of the TLA is 0×00002, i.e., it has a value of 2002::/16 when expressed as a 16-bit IPv6 address prefix. Hence, a 48-bit address prefix is created by the format prefix “001”, followed by the 13-bit TLA “0×00002”, followed by the assigned 32-bit IPv4 address; the remaining 80 bits of the 128-bit IPv6 address are available for a 16-bit Site-Level Aggregation Identifier (SLA ID), and a 64-bit Interface ID. Consequently, any node having at least one globally unique IPv4 address can have a globally unique IPv6 address, even if the node is combined with an IPv4 Network Address Translator (NAT).
Concerns arise in the event that an IPv6 node is coupled to a private IPv4 network having a Network Address Translator (NAT). NATs perform a Layer-3 translation of IP-Addresses, so that public Internet addresses map to private IP addresses, as described in detail by the Request for Comments 1918 (RFC 1918). This mapping has allowed enterprises to map a large number of private addresses to a limited number of public addresses, thus limiting the number of public addresses required by Internet users.
As described in RFC 3056, however, if an IPv6 node is coupled to an IPv4 network having a NAT, then the NAT box “must also contain a fully functional IPv6 router including the 6to4 mechanism” in order for the 6to4 proposal to still be operable in the IPv4 network having the NAT. However, the modification of existing NATs to include IPv6 routers to include the 6to4 mechanism may not be a practical solution.
Further, the IPv4 addresses of the 6to 4 protocol are assumed to be global public addresses. Hence, if an IPv6 node (i.e., a correspondent node) wants to communicate with a roaming mobile IPv6 node, the 6to4 address of the roaming mobile IPv6 node must be a global public address, not a private address.
One proposal for traversing a NAT by an IPv6 node using automatic tunneling is described in an IETF Draft by Huitema, entitled “Teredo: Tunneling IPv6 over UDP through NATs”, Sep. 17, 2002, available on the World Wide Web at the IETF website address “ietf.org/internet-drafts/draft-ietf-ngtrans-shipworm-08.txt”
Huitema suggests that IPv6 nodes located behind NATs can access “Teredo servers” and “Teredo relays” to learn their “global address” and to obtain connectivity, where clients, servers, and relays can be organized in “Teredo networks”. Huitema relies on a complex client server-based interaction between the client (i.e., the IPv6 node) behind the NAT in the private IPv4 address realm, and the Teredo server and Teredo relay on the opposite side of the NAT in the public IPv4 address realm. Hence, the communications between the IPv6 node on the private side of the NAT, and the Teredo server and the Teredo relay on the public side of the NAT, (via the NAT), require that the IPv6 node has a specified path to both the Teredo server and the Teredo relay gateway on the public side; hence, the IPv6 node needs to use the same IPv4 identifier (e.g., UDP port IP address that is being translated by the NAT).
This same IPv4 identifier for communications with the Teredo server and Teredo relay, however, is impossible with symmetric NATs because the Teredo server and Teredo relay each have a distinct corresponding IPv4 public address. In particular, symmetric NATs index their address translation tables not only by private IP address/private UDP port/ public IP address/ public UDP port of the packet output by a private IPv4 node, but also by the destination IP address and destination port specified by the packet and destined for the public IPv4 network. Hence, since the Teredo server and Teredo relay have distinct IP addresses, the symmetric NAT will not map the packets to the same private IP address/private UDP port used by the private IPv4 node. Hence, the Huitema solution cannot traverse a symmetric NAT.